- Email: [email protected]
Pregnancy outcome after large loop excision of the cervical transformation zone
September 1993 Am J Obstet Gynecol
Blomfield et al.
3. Barressi JA. Listeria monocytogenes: a cause of premature labor and neonatal sepsis. AM J OBSTET GYNECOL 1980; 136: 410-l. 4. Gembruch U, Niesen M, Hansmann M, Knopfle G. Listeriosis: a cause of non-immune hydrops fetalis. Prenat Diagn 1987;7:277-82. 5. Seeliger HPR, Bortolussi R. Listeriosis. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn. Philadelphia: W.B. Saunders, 1990:812-33. 6. Vawter GF. Perinatal listeriosis. Perspect Pediatr Pathol 1981 ;6: 153-66. 7. Steele PE, Jacobs DS. Listeria monocytogenes: macroabscesses of placenta. Obstet GynecoI1979;53:124-7. 8. Sarrut S, Alison F. Etude du placenta dans 21 cas de listeriose congenitale. Arch Fr Pediatr 1967;24:285-302.
9. Blanc WA. Pathology of the placenta, membranes, and umbilical cord in bacterial, fungal and viral infections in man. In: Naeye RL, Kissane JM, Kaufman N, eds. Perinatal diseases. Baltimore: Williams & Wilkins, 1981:67-132. 10. Benirschke K, Kaufman P. Pathology of the human placenta. 2nd ed. New York: Springer-Verlag, 1990:568-72. 11. Boucher M, Yonekura ML. Perinatal listeriosis (early onset): correlation of antenatal manifestations and neonatal outcome. Obstet Gynecol 1986;68:593-7. 12. Brown RC, Hopps HC. Staining of bacteria in tissue sections: a reliable gram stain method. Am J Clin Pathol 1973;60:234-40. 13. Garvey W, Fathi A, Bigelow F. Modified Steiner for the demonstration of spirochetes. J Histotech 1985;8:15-6.
Pregnancy outcome after large loop excision of the cervical transformation zone P.I. Blomfield, J. Buxton, J. Dunn, and D.M. Luesley, MD Birmingham, England OBJECTIVE: Our purpose was to determine whether large loop excision of the cervical transformation zone affects the outcome of pregnancy after 20 weeks' gestation. STUDY DESIGN: In a retrospective case control study 40 women who had undergone large loop excision of the cervical transformation zone and were subsequently delivered at Dudley Road Hospital were identified between January 1989 and January 1992. Eighty controls were identified and matched for age. parity, and ethnic group from women delivered immediately before and after index cases. Variables included maternal performance in labor and smoking habits and perinatal outcome. Maternal factors analyzed included pregnancy gestation, length of the first and second stages of labor, use of oxytocin, analgesia, mode of delivery, estimated blood loss, whether labor was spontaneous or induced. and if preterm spontaneous rupture of membranes occurred. Perinatal outcome measures included whether the infant was liveborn, fetal weight, the presence of fetal abnormalities, and admission to the neonatal unit. RESULTS: Women delivered after large loop excision of the cervical transformation zone had infants of significantly lower birth weight than did controls. They were also significantly more likely to have admitted to smoking on admission. CONCLUSION: Previous studies investigating pregnancy outcome after local destructive methods of treating cervical intraepithelial neoplasia have been generally reassuring. However, in this study women who were delivered after large loop excision of the cervical transformation zone had significantly smaller infants. Although this may be related to the characteristics of women who have cervical intraepithelial neoplasia (for example. their smoking habits), larger adequately controlled studies should be performed before colposcopists can be justified in adopting a liberal attitude to treating all women with abnormal smears. (AM J OSSTET GYNECOL 1993;169:620-5.)
Key words: Large loop excision of the cervical transformation zone, pregnancy outcome From the Academic Department of Obstetrics and Gynecology, Dudley Road Hospital. Received for publication November 9, 1992; accepted March 24, 1993. Reprint requests: P.I. Blomfield, 41 Union St., Armadale, Melbourne, Victoria 3143, Australia. Copyright © 1993 by Mosby-Year Book, Inc.
0002-9378/93 $1.00 + .20 6/1/47360
620
Since the turn of this century mortality from cervical cancer has been steadily declining. The impact of the whole machinery of cervical screening, colposcopic diagnosis, and treatment on crude mortality figures in the United Kingdom is currently not as marked as we might have hoped. The decline in mortality rates in this
Blomfield et al.
Volume 169, Number 3 Am J Obstet Gynecol
country in women > 35 years old since the 1960s is, however, interpreted by some to be a result of the beneficial effect of screening, and the increasing incidence seen in younger women may possibly have been larger in the absence of all this activity.! The trend toward more conservative modalities of treating cervical intraepithelial neoplasia, initially with cone biopsy and local destructive methods and now local excisional methods, has perhaps also been counterbalanced by a lower threshold for treatment. In spite of this, little is known about the side effects of treatment. Considerable psychologic and psychosexual morbidity associated with cot po scopic investigation and treatment has been documented. Little is known about the outcome of pregnancy after treatment. Large loop excision of the cervical transformation zone has been proposed as a safe and efficient method for conservatively treating cervical intraepithelial neoplasia. It is fast becoming one of the most popular modalities of treatment for this disorder. One of the main criticisms leveled at its use is that without the availability of improved predictors of disease a proportion of young women are overtreated for lower grade and possibly insignificant cervical lesions. One of the justifications for this criticism is the possible poorer reproductive outcome for women after treatment. Data on fertility rates after conservative management of cervical intraepithelial neoplasia are limited, but no major effect has been reported on subsequent fertility.2.4 Pregnancy complications after treatment of cervical intraepithelial neoplasia appear infrequent and seem confined mainly to patients who have undergone previous cone biopsy. Initial reports are based on patient series and comparison with expected outcome measures, and an increase in second-trimester losses and premature labor has been observed. 5 A relationship between cone depth and rate of midtrimester loss and premature labor was also demonstrated. 6 However, at the time most series were published there was liberal use of cervical cerclage, possibly influencing outcome. Very few case control studies have been published; those that have reveal no significant increase in early pregnancy loss, cesarean section, or premature delivery after cone biopsy.7 There also appears to be no adverse effect on pregnancy after local destructive therapies (i.e., laser vaporization or cryocautery).4-1O More than 1000 women have been treated at Dudley Road Hospital between 1988 and 1991. There are logistic difficulties in investigating fertility and early pregnancy loss rates after treatment. Pregnancy outcome after 20 weeks' gestation is more easily evaluated. To investigate whether prior treatment with large loop excision of the cervical transformation zone affected subsequent pregnancy performance, we carried out a retrospective case control study of pregnancy outcome after the twentieth week of gestation.
621
Material and methods Since February 1988 all patients treated in our colposcopy clinic with large loop excision of the cervical transformation zone have been recorded on a comprehensive computer data base. Labor performance and outcome has been computerized in the maternity unit since 1989. This has enabled the identification of 40 women who had undergone large loop excision of the cervical transformation zone and were subsequently delivered. For each index case two controls were selected matched for age, parity, and, because > 60% of the women delivering at Dudley Road Hospital are nonwhite, for ethnic group. Controls were selected from among women delivering immediately before and after the index case. Information was collected from the labor ward computer and from patient notes. Maternal factors included gestation, length of the first and second stages of labor, use of oxytocin, analgesia, mode of delivery, estimated blood loss, whether labor was spontaneous or induced, and if preterm spontaneous rupture of membranes occurred. Maternal smoking habits at booking were also recorded, as were the gestational age and birth weight of previous deliveries. Perinatal factors included whether the infant was live-born, fetal weight, presence of fetal abnormalities, and admission to the neonatal unit. Case-control comparison was carried out with the X2 test for grouped data and the Mann-Whitney U test for continuous variables.
Results Characteristics of women treated with large loop excision of the cervical transformation zone. Twentyone (52.5%) women delivered after undergoing large loop excision of the cervical transformation zone had been treated for high-grade cervical intraepithelial neoplasia (stages II and III), eight (20%) had stage I, nine (22.5%) had evidence of human papillomavirus infection, and two (5%) had adenocarcinoma in situ. The mean loop depth was 9.68 mm (range 7 to 16 mm), and the mean length of time from treatment to delivery was 19.5 months (range 6 to 43). Population definition. The median ages of both cases and controls were comparable, 26.0 years (interquantile range 23 to 31) and 25.0 years (interquantile range 23 to 30), respectively. Seventy percent of the study population were multiparous, 60% white, 22.5% Mro-Caribbean, 10% Asian, and 7.5% of other ethnic origins. Labor Intervention rates and delivery. Table I details the mode of delivery and the intervention rates in labor for both cases and controls. Only 7.5% of women delivered after undergoing large loop excision of the cervical transformation zone required a cesarean section, whereas
622
Blomfield et al.
September 1993 Am J Obstet Gynecol
Table I. Comparison of intervention rates and delivery after large loop excision of the cervical transformation zone and in controls Women delivered after large loop excision of cervical transformation zone (n = 40) Variable examined Mode of delivery Spontaneous vertex delivery Assisted breech delivery Forceps or ventouse delivery Emergency cesarean section Elective cesarean section Epidural rates Oxytocin in labor Induction rates
I
Control women (n = 80)
%
No.
34
85.0
63
2
5.0
No.
2
6 9 1
I
%
i'
Degrees of freedom
78.8
1.27
2
Significance
p=
0.26*
p= P=
0.41 1.0
1.3
2.5
4
5.0
5.0
9
11.3
2.5
3
3.8
15.0 22.5 2.5
17 18 4
21.3 22.5 5.0
0.67 0.001
t
t
*For comparison of abdominal versus vaginal delivery a X2 statistic was computed to compare all cesarean sections with all vaginal deliveries. tTest not available because of small numbers.
Table II. Comparison of labor performance Women delivered after large loop excision of cervical transformation zone Variable
Median
Length of first stage of labor (min) Length of second stage of labor (min)
307 19
I
Control women
Interquantile range
Median
184-488
310
10-35
15.1 % of controls were delivered by this method; however, this difference was not statistically significant. Labor performance. Excluding induction of labor and women delivered by cesarean section, no difference was seen in the length of the first or second stages of labor for the two groups (Table II). The mean length of both stages of labor was shorter for multiparous than for primiparous women, as expected (not detailed), but no case-control difference was seen. Oxytocin was used in 22.5% of women from both groups, and the overall time for which it was in use was directly comparable. Perinatal outcome. The mean birth weight of infants delivered of women after large loop excision of the cervical transformation zone was 2877 gm as opposed to the mean birth weight of infants delivered of control women, which was 3208 gm. This difference was significant when analyzed in a univariate analysis and became more powerfully so when only women who spontaneously labored or labored after premature rupture of
13
I
I nterquantile range
Mann- Whitney
210-540
p=
0.85
6-41
P=
0.11
membranes were included (i.e., women who underwent induction of labor or elective cesarean section were removed from the analysis, Table III). The mean gestational age at delivery was 37.4 weeks for women delivered after large loop excision of the cervical transformation zone and 38.7 weeks for controls. This difference was not significant and did not become so if only women spontaneously laboring were included. However, a higher number of women delivering after large loop excision of the cervical transformation zone went into spontaneous labor before 37 weeks, 15.8% compared with 10.9%. Overall, 17.5% of women treated by large loop excision of the cervical transformation zone had premature infants compared with 11.3% of control women (p = 0.34). The neonatal unit admission rate for infants delivered to women after large loop excision of the cervical transformation zone was twice that of controls, 10% compared with 5%. Two infants had minor congenital
Blomfield et al. 623
Volume 169, Number 3 Am J Obstet Gynecol
Table III. Perinatal outcome of women delivered after large loop excision of the cervical transformation zone compared with controls* Women delivered after large loop excision of cervical transformation zone Variable Gestational age (wk) <36 36-39 ~40
Birth weight (grn) 410-<2500 ~ 2500- < 3500 ~3500
I
Control women
I
I
Degrees of freedom
I
%
x!
7 33 33
9.6 45.2 45.2
0.36
2
0.84
5 45 23
6.8 61.7 31.5
6.77
2
0.03
%
No.
5 17 16
13.2 44.7 42.1
9 21 8
23.7 55.3 21.0
No.
Significance p value
*Women who spontaneous labored were included. Those delivered by elective cesarean section or who had been induced are excluded from the analysis.
Table IV. Comparison of obstetric histories between women delivered after large loop excision of the cervical transformation zone and controls
No. of previous early pregnancy losses Primigravid 0 ~1
Multiparous 0 1
Women delivered after large loop excision of cervical transformation zone No. 9 3
I
Control women
I
No.
75.0 25.0
19 5
79.2 20.8
<0.001
P=
1.0*
39 14 3
69.6 25.0 5.4
0.25
P=
0.62*
12 58 22
13.04 63.04 23.9
0.078
P=
0.96
18 64.3 17.8 5 17.9 ~2 5 Birth weight of previous infants born to multiparous women (gm) 13.5 <2500 5 64.8 2500-< 3500 24 ~3500 8 21.6
%
x!
%
Significance
*l analysis was carried out to compare women with no history of early pregnancy loss with those with one or more pregnancy losses. Yates' correction was used where the minimum expected value was < 5. abnormalities, one an extra digit, one a skin tag to an ear. Two women who were delivered after large loop excision of the cervical transformation zone had midtrimester losses, both at 22 weeks' gestation; one of these was the only twin pregnancy in the study. One further woman had an intrapartum death from abruptio placentae and congenitally acquired pneumonia at 38 weeks' gestation. One control woman also had midtrimester loss at 23 weeks' gestation, and a further control woman had an unexplained intrauterine death at 36 weeks' gestation. Although the pregnancy outcome for women delivered after large loop excision of the cervical transformation zone was poorer than for controls and the neonatal admission rate higher, meaningful interpretation cannot be made with such small numbers. Smoking habits and obstetric history. Also analyzed were the smoking habits of the population, in view of
the known association of cervical neoplasia with smoking. Unfortunately, this information was collected in a retrospective manner and the only available data related to whether women admitted to smoking in pregnancy at their first visit, normally at 16 weeks' gestation. A total of 62.4% delivered after large loop excision of the cervical transformation zone admitted to smoking compared with 37.6% of controls (X2 = 4.665, degrees of freedom = 1, P = 0.03). Obstetric histories of women delivered after large loop excision of the cervical transformation zone were compared with controls. Twenty-seven (67.5%) had never previously had a miscarriage or a therapeutic abortion, and 58 (72.5%) of control women were in a similar position. There was no increased frequency of a history of early pregnancy loss or termination of pregnancy among women delivered after large loop excision of the cervical transformation zone (Table IV). There
624 Blomfield et al.
was also no significant difference between the two groups in the birth weights and gestational age at delivery of previous infants born to multiparous women. The mean birth weight was 3030 gm, and gestation was 39.3 weeks for previous infants delivered of women after large loop excision of the cervical transformation zone, for control women 310 1 gm and 38.8 weeks, respectively.
Comment Treatment of women for cervical intraepithelial neoplasia by local destructive and excisional techniques has now become firmly established in this country. Many of these women are young and nulliparous. Studies investigating the outcome of pregnancy after treatment are considered to be reassuring." Large loop excision of the cervical transformation zone is becoming one of the most popular methods of treatment and has the added advantage of allowing histologic examination of the whole transformation zone. It is highly desirable that no treatment designed to eradicate cervical preinvasive disease should have adverse effects on future fertility or pregnancy outcome. We found that women previously treated by large loop excision of the cervical transformation zone delivered infants with significantly lower birth weights than controls and were more likely to be smokers. The study design did not control for smoking habits at the outset, controlling for parity, age, and ethnic group only. Only one previously published study investigating pregnancy outcome after conservative treatment of cervical intraepithelial neoplasia has included smoking habits.9 They too confirmed significantly higher numbers of women admitting to smoking in the treatment group. These findings are not unexpected considering the known association between smoking and cervical neoplasia. It is tempting to suggest that the obvious reason that treated women deliver lower birth weight infants is the known effect of smoking on pregnancy. We, however, found no direct association in our study population as a whole between women's smoking habits at first visit and birth weight (p = 0.91). We also found no difference in the previous birth weights of infants born to the multiparous women in the study. Both these observations do not support the hypothesis. Future study design should control for smoking and other known correlates of cervical intraepithelial neoplasia (e.g., social class) to disentangle these effects. The depth of excision at outpatient large loop excision of the cervical transformation zone procedures averages around 1 cm centrally; this probably represents more tissue destruction than with other local destructive techniques (e.g., laser and cryocautery). Depth and volume of destruction has in the past been
September 1993 Am J Obstet Gynecol
correlated with increased likelihood of premature labor and second-trimester loss.6 We found no increase in the rate of midtrimester losses after 20 weeks, and, although women delivered after large loop excision of the cervical transformation zone were more likely to delivery before 37 weeks (17.5% compared with 11.3%), this difference was not significant. In other case-control studies of pregnancy outcome after laser ablation of the cervical transformation zone, higher, although not statistically significant, rates of premature delivery in treated women were also observed. 9, 10 It is worrying that published studies consistently report higher rates of premature delivery. It is, however, possible that these observations are caused by other characteristics of women with cervical neoplasia rather than a direct effect of treatment. Alternatively, most studies include only small numbers of women, and, until large prospective matched case control studies are carried out, it is probably premature to be reassured by currently available data. Cases and controls appeared to behave similarly in labor; patients received a similar amount ofintervention, and the length of both stages of labor was the same for both groups. We found no evidence that primagravid women tended to be delivered more quickly after treatment, as reported by Buller and Jones. II There was no evidence of women after large loop excision of the cervical transformation zone having any worse past obstetric performance, and the similarity of the birth weights and gestational ages at delivery of previous infants born to multiparous women in the study does not support the argument suggesting that it is the characteristics of women who have cervical intraepithelial neoplasia that explain the case control difference. The logistics of studying pregnancy outcome after conservative treatment of cervical intraepithelial neoplasia have previously been difficult. With increasing in-house audit of colposcopy services and maternity units, identification of treated women should become much easier and should encourage further research. REFERENCES 1. Parkin DM, Nguyen-Dinh X, Day NE. The impact of screening on the incidence of cervical cancer in England and Wales. Br J Obstet Gynaecol 1985;92:150-7. 2. MacVicar J, Willocks J. The effect of diathermy conization of the cervix on subsequent fertility, pregnancy, and delivery. J Obstet Gynaecol Br Commonw 1968;75:355-6. 3. Green GH. Pregnancy following cervical carcinoma in situ. J Obstet Gynaecol Br Commonw 1966;73:897-902. 4. Hollyock VE, Chanen W, Wein R. Cervical function following treatment of intraepithelial neoplasia by electrocoagulation diathermy. Obstet Gynecol 1983;61:79-81. 5. Weber T, Obel E. Pregnancy complications following conization of the uterine cervix. Acta Obstet Gynecol Scand 1979;58:259-63. 6. Leiman G, Harrison NA, Rubin A. Pregnancy following
Barclay et al.
Volume 169, Number 3 Am J Obstet Gynecol
conization of the cervix: complications related to cone size.
AM J OBSTET GYNECOL 1980:36: 14-8.
7. Moinian M, Andersh B. Does cervix conization increase the risk of complication in subsequent pregnancy? Acta Obstet Gynecol Scand 1982;61:101-3. 8. Hammond RH, Edmonds DK. Does treatment for cervical intraepithelial neoplasia affect fertility and pregnancy? BMJ 1990;301:1344-5. 9. Saunders N, Fenton DW, Soutter WP, Brown VA, Sharp F,
Abernethy F. A case control study of the outcome of pregnancy following laser vapourization cone of the cervix. New York: Perinatology Press, 1985:121-32. 10. Anderson MC, Horwell DH, Broby Z. Outcome of pregnancy after laser vapourization. Colposc Gynecol Laser Surg 1984;2:35-42. 11. Buller RE, Jones HW. Pregnancy following cervical conization. AM J OBSTET GYNECOL 1982;142:506-12.
Interleukin-8 production by the human cervix Cameron G. Barclay, BSe; Janet E. Brennand, MB, ChB; Rodney W. Kelly, PhD,b and Andrew A. Calder, MD' Edinburgh, Scotland OBJECTIVES: Our purpose was (1) to determine whether the human cervix is capable of producing interleukin-8 in vitro and to examine the possibility of stimulating an increase in any such output and (2) to examine the concomitant production of prostaglandins. STUDY DESIGN: Cervical tissue was obtained from 48 women, 29 pregnant women undergoing surgical termination of pregnancy (20 of whom were treated with the prostaglandin analog Cervagem), 14 nonpregnant, premenopausal women, and three postmenopausal women. Explants were cultured and the medium was assayed for interleukin-8 and prostaglandin E2 • Analysis of variance and Newman-Keuls statistical tests were used. RESULTS: Significant quantities of interleukin-8 were produced by the tissue, and the data indicate that cervical explants from pregnant and nonpregnant women behave in a similar way when challenged by phorbol myristate acetate but that the postmenopausal cervix loses its capacity for interleukin-8 production. CONCLUSIONS: Human cervix is capable of producing large amounts of interleukin-8 in vitro, and it may be influenced by the steroid hormones. Thus interleukin-8 could be an excellent candidate for a prime role in neutrophil-mediated cervical ripening. (AM J OBSTET GYNECOL 1993;169:625-32.)
Key words: Interleukin-8, cervical ripening, neutrophils Cervical ripening is the process by which the cervix changes from a rigid, closed structure designed to retain the uterine contents to one that becomes soft and able to dilate to accommodate the passage of the fetus. The cervix consists mainly of connective tissue the majority of which, about 70%, is type I collagen. I Smooth muscle cells constitute about 10% of the cervix,2 whereas collagen-producing fibroblasts are the main cellular component. 3 Changes in the connective From the Department of Obstetrics and Gynaecology, University of Edinburgh, a and the Medical Research Council, Reproductive Biology Unit, Centre for Reproductive Biology.' Received for publication November 23, 1992; revised January 28, 1993; accepted March 8, 1993. Reprint requests: Rodney W. Kelly, PhD, Department of Obstetrics and Gynaecology, University of Edinburgh, Centre for Reproductive Biology, 37 Chalmers St., Edinburgh, Scotland EH3 9EW. Copyright © 1993 by Mosby-Year Book, Inc. 0002-9378/93 $1.00 + .20 6/1/46929
tissues occur throughout pregnancy, I but this remodeling of the collagen does not greatly alter the rigidity of the cervix. Only in the very last stages can its mechanical strength be permitted to change in such a dramatic manner as to permit fetal expulsion. It has been suggested that this process is similar to an inflammatory reaction,4 and Junqueira et al. 3 and Rath et al. 5 have shown that neutrophil invasion of the cervix occurs in labor. Others have proposed that the breakdown of the collagen matrix in the final stages of pregnancy is caused by collagenase and elastase produced not only by leukocytes, as would be found in an inflammatory reaction, but also by the fibroblasts of the cervix, which are known to produce collagenase. I Collagenase produced by granulocytes, such as neutrophils, selectively degTades type I collagen,6 whereas collagenase produced by fibroblasts degTades type I and type III collagen equally. Because type III collagen
625
Blomfield et al.
3. Barressi JA. Listeria monocytogenes: a cause of premature labor and neonatal sepsis. AM J OBSTET GYNECOL 1980; 136: 410-l. 4. Gembruch U, Niesen M, Hansmann M, Knopfle G. Listeriosis: a cause of non-immune hydrops fetalis. Prenat Diagn 1987;7:277-82. 5. Seeliger HPR, Bortolussi R. Listeriosis. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn. Philadelphia: W.B. Saunders, 1990:812-33. 6. Vawter GF. Perinatal listeriosis. Perspect Pediatr Pathol 1981 ;6: 153-66. 7. Steele PE, Jacobs DS. Listeria monocytogenes: macroabscesses of placenta. Obstet GynecoI1979;53:124-7. 8. Sarrut S, Alison F. Etude du placenta dans 21 cas de listeriose congenitale. Arch Fr Pediatr 1967;24:285-302.
9. Blanc WA. Pathology of the placenta, membranes, and umbilical cord in bacterial, fungal and viral infections in man. In: Naeye RL, Kissane JM, Kaufman N, eds. Perinatal diseases. Baltimore: Williams & Wilkins, 1981:67-132. 10. Benirschke K, Kaufman P. Pathology of the human placenta. 2nd ed. New York: Springer-Verlag, 1990:568-72. 11. Boucher M, Yonekura ML. Perinatal listeriosis (early onset): correlation of antenatal manifestations and neonatal outcome. Obstet Gynecol 1986;68:593-7. 12. Brown RC, Hopps HC. Staining of bacteria in tissue sections: a reliable gram stain method. Am J Clin Pathol 1973;60:234-40. 13. Garvey W, Fathi A, Bigelow F. Modified Steiner for the demonstration of spirochetes. J Histotech 1985;8:15-6.
Pregnancy outcome after large loop excision of the cervical transformation zone P.I. Blomfield, J. Buxton, J. Dunn, and D.M. Luesley, MD Birmingham, England OBJECTIVE: Our purpose was to determine whether large loop excision of the cervical transformation zone affects the outcome of pregnancy after 20 weeks' gestation. STUDY DESIGN: In a retrospective case control study 40 women who had undergone large loop excision of the cervical transformation zone and were subsequently delivered at Dudley Road Hospital were identified between January 1989 and January 1992. Eighty controls were identified and matched for age. parity, and ethnic group from women delivered immediately before and after index cases. Variables included maternal performance in labor and smoking habits and perinatal outcome. Maternal factors analyzed included pregnancy gestation, length of the first and second stages of labor, use of oxytocin, analgesia, mode of delivery, estimated blood loss, whether labor was spontaneous or induced. and if preterm spontaneous rupture of membranes occurred. Perinatal outcome measures included whether the infant was liveborn, fetal weight, the presence of fetal abnormalities, and admission to the neonatal unit. RESULTS: Women delivered after large loop excision of the cervical transformation zone had infants of significantly lower birth weight than did controls. They were also significantly more likely to have admitted to smoking on admission. CONCLUSION: Previous studies investigating pregnancy outcome after local destructive methods of treating cervical intraepithelial neoplasia have been generally reassuring. However, in this study women who were delivered after large loop excision of the cervical transformation zone had significantly smaller infants. Although this may be related to the characteristics of women who have cervical intraepithelial neoplasia (for example. their smoking habits), larger adequately controlled studies should be performed before colposcopists can be justified in adopting a liberal attitude to treating all women with abnormal smears. (AM J OSSTET GYNECOL 1993;169:620-5.)
Key words: Large loop excision of the cervical transformation zone, pregnancy outcome From the Academic Department of Obstetrics and Gynecology, Dudley Road Hospital. Received for publication November 9, 1992; accepted March 24, 1993. Reprint requests: P.I. Blomfield, 41 Union St., Armadale, Melbourne, Victoria 3143, Australia. Copyright © 1993 by Mosby-Year Book, Inc.
0002-9378/93 $1.00 + .20 6/1/47360
620
Since the turn of this century mortality from cervical cancer has been steadily declining. The impact of the whole machinery of cervical screening, colposcopic diagnosis, and treatment on crude mortality figures in the United Kingdom is currently not as marked as we might have hoped. The decline in mortality rates in this
Blomfield et al.
Volume 169, Number 3 Am J Obstet Gynecol
country in women > 35 years old since the 1960s is, however, interpreted by some to be a result of the beneficial effect of screening, and the increasing incidence seen in younger women may possibly have been larger in the absence of all this activity.! The trend toward more conservative modalities of treating cervical intraepithelial neoplasia, initially with cone biopsy and local destructive methods and now local excisional methods, has perhaps also been counterbalanced by a lower threshold for treatment. In spite of this, little is known about the side effects of treatment. Considerable psychologic and psychosexual morbidity associated with cot po scopic investigation and treatment has been documented. Little is known about the outcome of pregnancy after treatment. Large loop excision of the cervical transformation zone has been proposed as a safe and efficient method for conservatively treating cervical intraepithelial neoplasia. It is fast becoming one of the most popular modalities of treatment for this disorder. One of the main criticisms leveled at its use is that without the availability of improved predictors of disease a proportion of young women are overtreated for lower grade and possibly insignificant cervical lesions. One of the justifications for this criticism is the possible poorer reproductive outcome for women after treatment. Data on fertility rates after conservative management of cervical intraepithelial neoplasia are limited, but no major effect has been reported on subsequent fertility.2.4 Pregnancy complications after treatment of cervical intraepithelial neoplasia appear infrequent and seem confined mainly to patients who have undergone previous cone biopsy. Initial reports are based on patient series and comparison with expected outcome measures, and an increase in second-trimester losses and premature labor has been observed. 5 A relationship between cone depth and rate of midtrimester loss and premature labor was also demonstrated. 6 However, at the time most series were published there was liberal use of cervical cerclage, possibly influencing outcome. Very few case control studies have been published; those that have reveal no significant increase in early pregnancy loss, cesarean section, or premature delivery after cone biopsy.7 There also appears to be no adverse effect on pregnancy after local destructive therapies (i.e., laser vaporization or cryocautery).4-1O More than 1000 women have been treated at Dudley Road Hospital between 1988 and 1991. There are logistic difficulties in investigating fertility and early pregnancy loss rates after treatment. Pregnancy outcome after 20 weeks' gestation is more easily evaluated. To investigate whether prior treatment with large loop excision of the cervical transformation zone affected subsequent pregnancy performance, we carried out a retrospective case control study of pregnancy outcome after the twentieth week of gestation.
621
Material and methods Since February 1988 all patients treated in our colposcopy clinic with large loop excision of the cervical transformation zone have been recorded on a comprehensive computer data base. Labor performance and outcome has been computerized in the maternity unit since 1989. This has enabled the identification of 40 women who had undergone large loop excision of the cervical transformation zone and were subsequently delivered. For each index case two controls were selected matched for age, parity, and, because > 60% of the women delivering at Dudley Road Hospital are nonwhite, for ethnic group. Controls were selected from among women delivering immediately before and after the index case. Information was collected from the labor ward computer and from patient notes. Maternal factors included gestation, length of the first and second stages of labor, use of oxytocin, analgesia, mode of delivery, estimated blood loss, whether labor was spontaneous or induced, and if preterm spontaneous rupture of membranes occurred. Maternal smoking habits at booking were also recorded, as were the gestational age and birth weight of previous deliveries. Perinatal factors included whether the infant was live-born, fetal weight, presence of fetal abnormalities, and admission to the neonatal unit. Case-control comparison was carried out with the X2 test for grouped data and the Mann-Whitney U test for continuous variables.
Results Characteristics of women treated with large loop excision of the cervical transformation zone. Twentyone (52.5%) women delivered after undergoing large loop excision of the cervical transformation zone had been treated for high-grade cervical intraepithelial neoplasia (stages II and III), eight (20%) had stage I, nine (22.5%) had evidence of human papillomavirus infection, and two (5%) had adenocarcinoma in situ. The mean loop depth was 9.68 mm (range 7 to 16 mm), and the mean length of time from treatment to delivery was 19.5 months (range 6 to 43). Population definition. The median ages of both cases and controls were comparable, 26.0 years (interquantile range 23 to 31) and 25.0 years (interquantile range 23 to 30), respectively. Seventy percent of the study population were multiparous, 60% white, 22.5% Mro-Caribbean, 10% Asian, and 7.5% of other ethnic origins. Labor Intervention rates and delivery. Table I details the mode of delivery and the intervention rates in labor for both cases and controls. Only 7.5% of women delivered after undergoing large loop excision of the cervical transformation zone required a cesarean section, whereas
622
Blomfield et al.
September 1993 Am J Obstet Gynecol
Table I. Comparison of intervention rates and delivery after large loop excision of the cervical transformation zone and in controls Women delivered after large loop excision of cervical transformation zone (n = 40) Variable examined Mode of delivery Spontaneous vertex delivery Assisted breech delivery Forceps or ventouse delivery Emergency cesarean section Elective cesarean section Epidural rates Oxytocin in labor Induction rates
I
Control women (n = 80)
%
No.
34
85.0
63
2
5.0
No.
2
6 9 1
I
%
i'
Degrees of freedom
78.8
1.27
2
Significance
p=
0.26*
p= P=
0.41 1.0
1.3
2.5
4
5.0
5.0
9
11.3
2.5
3
3.8
15.0 22.5 2.5
17 18 4
21.3 22.5 5.0
0.67 0.001
t
t
*For comparison of abdominal versus vaginal delivery a X2 statistic was computed to compare all cesarean sections with all vaginal deliveries. tTest not available because of small numbers.
Table II. Comparison of labor performance Women delivered after large loop excision of cervical transformation zone Variable
Median
Length of first stage of labor (min) Length of second stage of labor (min)
307 19
I
Control women
Interquantile range
Median
184-488
310
10-35
15.1 % of controls were delivered by this method; however, this difference was not statistically significant. Labor performance. Excluding induction of labor and women delivered by cesarean section, no difference was seen in the length of the first or second stages of labor for the two groups (Table II). The mean length of both stages of labor was shorter for multiparous than for primiparous women, as expected (not detailed), but no case-control difference was seen. Oxytocin was used in 22.5% of women from both groups, and the overall time for which it was in use was directly comparable. Perinatal outcome. The mean birth weight of infants delivered of women after large loop excision of the cervical transformation zone was 2877 gm as opposed to the mean birth weight of infants delivered of control women, which was 3208 gm. This difference was significant when analyzed in a univariate analysis and became more powerfully so when only women who spontaneously labored or labored after premature rupture of
13
I
I nterquantile range
Mann- Whitney
210-540
p=
0.85
6-41
P=
0.11
membranes were included (i.e., women who underwent induction of labor or elective cesarean section were removed from the analysis, Table III). The mean gestational age at delivery was 37.4 weeks for women delivered after large loop excision of the cervical transformation zone and 38.7 weeks for controls. This difference was not significant and did not become so if only women spontaneously laboring were included. However, a higher number of women delivering after large loop excision of the cervical transformation zone went into spontaneous labor before 37 weeks, 15.8% compared with 10.9%. Overall, 17.5% of women treated by large loop excision of the cervical transformation zone had premature infants compared with 11.3% of control women (p = 0.34). The neonatal unit admission rate for infants delivered to women after large loop excision of the cervical transformation zone was twice that of controls, 10% compared with 5%. Two infants had minor congenital
Blomfield et al. 623
Volume 169, Number 3 Am J Obstet Gynecol
Table III. Perinatal outcome of women delivered after large loop excision of the cervical transformation zone compared with controls* Women delivered after large loop excision of cervical transformation zone Variable Gestational age (wk) <36 36-39 ~40
Birth weight (grn) 410-<2500 ~ 2500- < 3500 ~3500
I
Control women
I
I
Degrees of freedom
I
%
x!
7 33 33
9.6 45.2 45.2
0.36
2
0.84
5 45 23
6.8 61.7 31.5
6.77
2
0.03
%
No.
5 17 16
13.2 44.7 42.1
9 21 8
23.7 55.3 21.0
No.
Significance p value
*Women who spontaneous labored were included. Those delivered by elective cesarean section or who had been induced are excluded from the analysis.
Table IV. Comparison of obstetric histories between women delivered after large loop excision of the cervical transformation zone and controls
No. of previous early pregnancy losses Primigravid 0 ~1
Multiparous 0 1
Women delivered after large loop excision of cervical transformation zone No. 9 3
I
Control women
I
No.
75.0 25.0
19 5
79.2 20.8
<0.001
P=
1.0*
39 14 3
69.6 25.0 5.4
0.25
P=
0.62*
12 58 22
13.04 63.04 23.9
0.078
P=
0.96
18 64.3 17.8 5 17.9 ~2 5 Birth weight of previous infants born to multiparous women (gm) 13.5 <2500 5 64.8 2500-< 3500 24 ~3500 8 21.6
%
x!
%
Significance
*l analysis was carried out to compare women with no history of early pregnancy loss with those with one or more pregnancy losses. Yates' correction was used where the minimum expected value was < 5. abnormalities, one an extra digit, one a skin tag to an ear. Two women who were delivered after large loop excision of the cervical transformation zone had midtrimester losses, both at 22 weeks' gestation; one of these was the only twin pregnancy in the study. One further woman had an intrapartum death from abruptio placentae and congenitally acquired pneumonia at 38 weeks' gestation. One control woman also had midtrimester loss at 23 weeks' gestation, and a further control woman had an unexplained intrauterine death at 36 weeks' gestation. Although the pregnancy outcome for women delivered after large loop excision of the cervical transformation zone was poorer than for controls and the neonatal admission rate higher, meaningful interpretation cannot be made with such small numbers. Smoking habits and obstetric history. Also analyzed were the smoking habits of the population, in view of
the known association of cervical neoplasia with smoking. Unfortunately, this information was collected in a retrospective manner and the only available data related to whether women admitted to smoking in pregnancy at their first visit, normally at 16 weeks' gestation. A total of 62.4% delivered after large loop excision of the cervical transformation zone admitted to smoking compared with 37.6% of controls (X2 = 4.665, degrees of freedom = 1, P = 0.03). Obstetric histories of women delivered after large loop excision of the cervical transformation zone were compared with controls. Twenty-seven (67.5%) had never previously had a miscarriage or a therapeutic abortion, and 58 (72.5%) of control women were in a similar position. There was no increased frequency of a history of early pregnancy loss or termination of pregnancy among women delivered after large loop excision of the cervical transformation zone (Table IV). There
624 Blomfield et al.
was also no significant difference between the two groups in the birth weights and gestational age at delivery of previous infants born to multiparous women. The mean birth weight was 3030 gm, and gestation was 39.3 weeks for previous infants delivered of women after large loop excision of the cervical transformation zone, for control women 310 1 gm and 38.8 weeks, respectively.
Comment Treatment of women for cervical intraepithelial neoplasia by local destructive and excisional techniques has now become firmly established in this country. Many of these women are young and nulliparous. Studies investigating the outcome of pregnancy after treatment are considered to be reassuring." Large loop excision of the cervical transformation zone is becoming one of the most popular methods of treatment and has the added advantage of allowing histologic examination of the whole transformation zone. It is highly desirable that no treatment designed to eradicate cervical preinvasive disease should have adverse effects on future fertility or pregnancy outcome. We found that women previously treated by large loop excision of the cervical transformation zone delivered infants with significantly lower birth weights than controls and were more likely to be smokers. The study design did not control for smoking habits at the outset, controlling for parity, age, and ethnic group only. Only one previously published study investigating pregnancy outcome after conservative treatment of cervical intraepithelial neoplasia has included smoking habits.9 They too confirmed significantly higher numbers of women admitting to smoking in the treatment group. These findings are not unexpected considering the known association between smoking and cervical neoplasia. It is tempting to suggest that the obvious reason that treated women deliver lower birth weight infants is the known effect of smoking on pregnancy. We, however, found no direct association in our study population as a whole between women's smoking habits at first visit and birth weight (p = 0.91). We also found no difference in the previous birth weights of infants born to the multiparous women in the study. Both these observations do not support the hypothesis. Future study design should control for smoking and other known correlates of cervical intraepithelial neoplasia (e.g., social class) to disentangle these effects. The depth of excision at outpatient large loop excision of the cervical transformation zone procedures averages around 1 cm centrally; this probably represents more tissue destruction than with other local destructive techniques (e.g., laser and cryocautery). Depth and volume of destruction has in the past been
September 1993 Am J Obstet Gynecol
correlated with increased likelihood of premature labor and second-trimester loss.6 We found no increase in the rate of midtrimester losses after 20 weeks, and, although women delivered after large loop excision of the cervical transformation zone were more likely to delivery before 37 weeks (17.5% compared with 11.3%), this difference was not significant. In other case-control studies of pregnancy outcome after laser ablation of the cervical transformation zone, higher, although not statistically significant, rates of premature delivery in treated women were also observed. 9, 10 It is worrying that published studies consistently report higher rates of premature delivery. It is, however, possible that these observations are caused by other characteristics of women with cervical neoplasia rather than a direct effect of treatment. Alternatively, most studies include only small numbers of women, and, until large prospective matched case control studies are carried out, it is probably premature to be reassured by currently available data. Cases and controls appeared to behave similarly in labor; patients received a similar amount ofintervention, and the length of both stages of labor was the same for both groups. We found no evidence that primagravid women tended to be delivered more quickly after treatment, as reported by Buller and Jones. II There was no evidence of women after large loop excision of the cervical transformation zone having any worse past obstetric performance, and the similarity of the birth weights and gestational ages at delivery of previous infants born to multiparous women in the study does not support the argument suggesting that it is the characteristics of women who have cervical intraepithelial neoplasia that explain the case control difference. The logistics of studying pregnancy outcome after conservative treatment of cervical intraepithelial neoplasia have previously been difficult. With increasing in-house audit of colposcopy services and maternity units, identification of treated women should become much easier and should encourage further research. REFERENCES 1. Parkin DM, Nguyen-Dinh X, Day NE. The impact of screening on the incidence of cervical cancer in England and Wales. Br J Obstet Gynaecol 1985;92:150-7. 2. MacVicar J, Willocks J. The effect of diathermy conization of the cervix on subsequent fertility, pregnancy, and delivery. J Obstet Gynaecol Br Commonw 1968;75:355-6. 3. Green GH. Pregnancy following cervical carcinoma in situ. J Obstet Gynaecol Br Commonw 1966;73:897-902. 4. Hollyock VE, Chanen W, Wein R. Cervical function following treatment of intraepithelial neoplasia by electrocoagulation diathermy. Obstet Gynecol 1983;61:79-81. 5. Weber T, Obel E. Pregnancy complications following conization of the uterine cervix. Acta Obstet Gynecol Scand 1979;58:259-63. 6. Leiman G, Harrison NA, Rubin A. Pregnancy following
Barclay et al.
Volume 169, Number 3 Am J Obstet Gynecol
conization of the cervix: complications related to cone size.
AM J OBSTET GYNECOL 1980:36: 14-8.
7. Moinian M, Andersh B. Does cervix conization increase the risk of complication in subsequent pregnancy? Acta Obstet Gynecol Scand 1982;61:101-3. 8. Hammond RH, Edmonds DK. Does treatment for cervical intraepithelial neoplasia affect fertility and pregnancy? BMJ 1990;301:1344-5. 9. Saunders N, Fenton DW, Soutter WP, Brown VA, Sharp F,
Abernethy F. A case control study of the outcome of pregnancy following laser vapourization cone of the cervix. New York: Perinatology Press, 1985:121-32. 10. Anderson MC, Horwell DH, Broby Z. Outcome of pregnancy after laser vapourization. Colposc Gynecol Laser Surg 1984;2:35-42. 11. Buller RE, Jones HW. Pregnancy following cervical conization. AM J OBSTET GYNECOL 1982;142:506-12.
Interleukin-8 production by the human cervix Cameron G. Barclay, BSe; Janet E. Brennand, MB, ChB; Rodney W. Kelly, PhD,b and Andrew A. Calder, MD' Edinburgh, Scotland OBJECTIVES: Our purpose was (1) to determine whether the human cervix is capable of producing interleukin-8 in vitro and to examine the possibility of stimulating an increase in any such output and (2) to examine the concomitant production of prostaglandins. STUDY DESIGN: Cervical tissue was obtained from 48 women, 29 pregnant women undergoing surgical termination of pregnancy (20 of whom were treated with the prostaglandin analog Cervagem), 14 nonpregnant, premenopausal women, and three postmenopausal women. Explants were cultured and the medium was assayed for interleukin-8 and prostaglandin E2 • Analysis of variance and Newman-Keuls statistical tests were used. RESULTS: Significant quantities of interleukin-8 were produced by the tissue, and the data indicate that cervical explants from pregnant and nonpregnant women behave in a similar way when challenged by phorbol myristate acetate but that the postmenopausal cervix loses its capacity for interleukin-8 production. CONCLUSIONS: Human cervix is capable of producing large amounts of interleukin-8 in vitro, and it may be influenced by the steroid hormones. Thus interleukin-8 could be an excellent candidate for a prime role in neutrophil-mediated cervical ripening. (AM J OBSTET GYNECOL 1993;169:625-32.)
Key words: Interleukin-8, cervical ripening, neutrophils Cervical ripening is the process by which the cervix changes from a rigid, closed structure designed to retain the uterine contents to one that becomes soft and able to dilate to accommodate the passage of the fetus. The cervix consists mainly of connective tissue the majority of which, about 70%, is type I collagen. I Smooth muscle cells constitute about 10% of the cervix,2 whereas collagen-producing fibroblasts are the main cellular component. 3 Changes in the connective From the Department of Obstetrics and Gynaecology, University of Edinburgh, a and the Medical Research Council, Reproductive Biology Unit, Centre for Reproductive Biology.' Received for publication November 23, 1992; revised January 28, 1993; accepted March 8, 1993. Reprint requests: Rodney W. Kelly, PhD, Department of Obstetrics and Gynaecology, University of Edinburgh, Centre for Reproductive Biology, 37 Chalmers St., Edinburgh, Scotland EH3 9EW. Copyright © 1993 by Mosby-Year Book, Inc. 0002-9378/93 $1.00 + .20 6/1/46929
tissues occur throughout pregnancy, I but this remodeling of the collagen does not greatly alter the rigidity of the cervix. Only in the very last stages can its mechanical strength be permitted to change in such a dramatic manner as to permit fetal expulsion. It has been suggested that this process is similar to an inflammatory reaction,4 and Junqueira et al. 3 and Rath et al. 5 have shown that neutrophil invasion of the cervix occurs in labor. Others have proposed that the breakdown of the collagen matrix in the final stages of pregnancy is caused by collagenase and elastase produced not only by leukocytes, as would be found in an inflammatory reaction, but also by the fibroblasts of the cervix, which are known to produce collagenase. I Collagenase produced by granulocytes, such as neutrophils, selectively degTades type I collagen,6 whereas collagenase produced by fibroblasts degTades type I and type III collagen equally. Because type III collagen
625